buffer

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill instructs the agent to read Emacs buffers and "return the full text to the user" verbatim (via emacsclient), which will cause any secrets present in those buffers (API keys, tokens, passwords) to be output directly by the LLM.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The skill reads arbitrary Emacs buffer contents at runtime by treating buffer-name references in outsider-authored user prompts as requests, then running emacsclient --eval to call agent-skill-buffer which returns the buffer text into the conversation.